IL-21 Promotes the Pathologic Immune Response to Pneumovirus Infection

Interleukin-21 is a pleiotropic four α-helical bundle type I cytokine that is produced primarily by CD4⁺ T cell populations, including T follicular helper (Tfh) cells, Th17 populations, and NKT cells, and acts on a broad range of target cells including B cells, T cells, NK cells, dendritic cells (DCs), macrophages, and epithelial cells (1). The functional IL-21R is a heterodimer of an IL-21–specific protein, IL-21R, and the common cytokine receptor γ-chain, γ_c, which is shared by the receptors for IL-2, IL-4, IL-7, IL-9, and IL-15 (2) and is encoded by the gene that is mutated in humans with X-linked SCID (3). IL-21 functions as a T cell comitogen and can also cooperate with IL-7 and IL-15 to drive expansion of CD8⁺ T cells (4). Within the B cell lineage, it is a critical regulator of Ig production, contributes to Ig class switch, and drives terminal B cell differentiation to plasma cells via its regulation of the PRDM1 gene, which encodes BLIMP1 (5, 6). Interestingly, IL-21 is also proapoptotic for B cells (5, 7, 8) and NK cells (9) and can suppress DC function (10). IL-21 plays a role in both CD8⁺ T cell and NK cell function related to anti-tumor responses, and IL-21 exhibits potent anticancer activity in a number of animal models (1). Although IL-21 exerts positive effects on differentiation and function, it also has inhibitory effects on immune responses based on its role in B cell apoptosis, its induction of IL-10 in T and B cells leading to immunosuppression (11), and its suppressive effects on DC function (10). IL-21 also promotes the development and expansion of the inflammatory Th17 subset of CD4⁺ T cells (12–14) as well as Tfh cells (15).

IL-21 can promote inflammation in a variety of disease-related settings. For example, in the NOD mouse model of type 1 diabetes, IL-21 signaling promotes pancreatic islet inflammation and destruction, as shown by the lack of disease when mice are crossed to the Il21r^−/− background (16, 17). Analogously, IL-21 also promotes autoimmune systemic lupus erythematosus related to its role in T cell activation, with diminished disease development in two different mouse models when IL-21 signaling is blocked (18, 19). Moreover, in the experimental autoimmune encephalitis mouse model of multiple sclerosis, IL-21 promotes the differentiation and expansion of IL-17–producing CD4⁺ T cells, and the absence of IL-21 signaling has been associated with reduced severity of disease (12, 13).

Despite the large body of information known as being related to the biology of IL-21, relatively limited information is available regarding its role following viral infection. In HIV-infected individuals, IL-21–expressing CD4⁺ T cells were present, and these numbers decreased as viral levels declined (20). Additionally, IL-21–producing HIV-specific CD8⁺ T cells were induced following HIV infection, and interestingly, the presence of these cells correlated better with HIV-1 control than did IL-21–producing CD4⁺ T cells (21). However, these findings were only correlative, and a full understanding of the role of IL-21 in host defense to HIV requires additional investigation. IL-21 is also elevated during lymphocytic choriomeningitis virus infection of mice, and experiments with Il21r^−/− mice revealed a role for IL-21 in controlling CD8⁺ T cell function during this infection (22–24). However, the role of IL-21 in the pathogenesis of acute respiratory viruses has not yet been explored. In this study, we focus on pneumonia virus of mice (PVM), a negative-strand RNA virus that is in the same family (Paramyxoviridae) and genus (Pneumovirus) as human respiratory syncytial virus (RSV) (25, 26). PVM infection can result in disease in mice similar to the more severe forms of RSV infection, with viral replication in respiratory epithelial cells that is associated with the infiltration of neutrophils and lymphocytes and the production of IFN-γ and the chemokines CCL2, CCL3, CXCL1, and CXCL10. In this study, we examined the role of IL-21 in both the innate and adaptive responses to PVM infection and unexpectedly found that mice lacking Il21r^−/− exhibited prolonged survival to PVM. Moreover, treatment of wild-type (WT) mice with a soluble IL-21R–Fc fusion protein was protective, indicating a role for IL-21 in promoting the pathogenic effects of this respiratory virus.

WT C57BL/6 mice were obtained from The Jackson Laboratory. Il21r^−/− mice (6) were analyzed at 8–12 wk of age. Mice expressing the Il2-EmGFP/Il21-mCherry recombineered Bac reporter transgene have been described (27). Transgenic mice expressing the human IL-21 cDNA under the control of the H-2K^b promoter and IgM enhancer have been described (5). All experiments were performed under protocols approved by the National Heart, Lung, and Blood Institute and National Institute of Allergy and Infectious Diseases Animal Care and Use Committees and followed National Institutes of Health guidelines for use of animals in intramural research.

Virus stocks (PVM strain J3666) were prepared as previously described (28). Mice were anesthetized briefly via inhalation of 20% halothane and inoculated intranasally with 12–60 PFU PVM in 50–100 μl PBS. IL-21R/Fc fusion protein was obtained from R&D Systems. Human Fc Cγ1 control protein was obtained from BioXCell.

Lungs were inflated before excision, fixed in 10% formalin, and embedded in paraffin, 5-μm thick sections were cut, and slides were stained with H&E. Clinical scores for morphology, inflammation, and edema were evaluated.

Lungs were inflated intratracheally with 1 ml cold 0.1% BSA in PBS, and recovered fluid was used for ELISA and to prepare bronchoalveolar lavage (BAL) cells. Lung tissue was minced into small pieces using a razor blade and digested in a solution containing 0.5 mg/ml Liberase (Roche) and 0.5 mg/ml DNase I (Sigma-Aldrich) in serum-free RPMI 1640 for 30 min at 37°C. Digested tissue was then pushed through a cell strainer with a syringe. Cells were centrifuged, and RBCs were lysed with ACK, followed by two washes with complete RPMI 1640.

Single-cell suspensions from BAL fluid or lung tissue were surface-stained in FACS buffer (PBS containing 0.5% BSA and 0.02% azide) using Abs from BD Biosciences. For intracellular staining of cytokines, cells were activated with PMA (10 ng/ml) and ionomycin (1 μM; Sigma-Aldrich) for 4 h in the presence of Golgi Plug (BD Biosciences). Cells were first surface-stained with anti-CD8, fixed, and permeabilized with CytoFix/CytoPerm (BD Biosciences). Data were acquired with either a BD FACS Canto II or LSR II (BD Biosciences) and analyzed with FlowJo software (Tree Star).

Splenic DCs were isolated by collagenase digestion followed by positive selection with pan-DC microbeads (Miltenyi Biotec). Splenic DCs were seeded at 2 × 10⁶/sample in 24-well plates, rested at 37°C for 1 h, and stimulated with cytokines. Supernatants were collected, and IL-6 levels were measured by ELISA (BD Biosciences).

RNA was extracted from lung tissue by homogenization in TRIzol (Invitrogen) followed by RNA cleanup with the RNeasy kit (Qiagen). RNA was reverse-transcribed using Omniscript (Qiagen). PCR reactions to quantitatively detect cytokine and chemokine RNAs used probe sets from Applied Biosystems and the 7900HT Sequence Detection System. Relative levels of PVM SH gene expression (29) were determined by quantitative RT-PCR, with normalization to Rpl7 expression.

Lung tissue was homogenized in cold PBS, and tissue debris was removed by centrifugation (14,000 rpm for 10 min). Levels of CXCL1 in equal volumes of BAL fluid and equal amounts of protein from lung homogenates were measured using a kit (Antigenix).

Two-tailed paired t tests were performed using Prism 4.0 (GraphPad). In the figures, statistical significance is indicated by *p < 0.05, **p < 0.01, and ***p < 0.001.

To determine whether IL-21 is expressed in the lung in response to PVM infection, WT mice were inoculated intranasally with PVM virions, and IL-21 mRNA expression was measured in the lung at time points during the 10 d following infection. Expression of the PVM SH gene as an indicator of virus burden (29) was greatest at day 6 after inoculation (Fig. 1A). As expected, infected mice developed pneumonia and began to succumb to infection starting at day 7, and SH gene expression declined, with it no longer being detected among survivors at day 10 (Fig. 1A). Il21 mRNA was detected as early as day 5, with a peak at day 6, and levels declined thereafter (Fig. 1B). Interestingly, at day 10, virus was no longer detected in lung tissue, but Il21 mRNA could be detected in the surviving mice.

To identify IL-21–producing cells, we used transgenic reporter mice in which the mCherry reporter is inserted at the translation start site of the Il21 gene in a bacterial artificial chromosome clone by recombineering technology. These mice were inoculated with PVM, and single-cell suspensions from lung tissue and lymphoid organs were examined ex vivo by flow cytometry. mCherry expression was detected in ∼5% of CD4⁺ T cells in lung tissue of uninfected mice, with expression increasing to ∼10–15% at day 6 after inoculation (Fig. 1C, 1D), which corresponded to peak IL-21 mRNA expression (Fig. 1B), whereas mCherry expression was not detected in NK1.1⁺ cells (NK cells or NKT cells), γδ T cells, or CD8⁺ T cells in the lung (Supplemental Fig. 1). In addition to this acute response in the lung, a systemic immune response also developed, as indicated by the increase in mCherry expression from ∼1.9% of splenic CD4⁺ T cells in uninfected reporter mice (TG Ctrl) to ∼5.9% at day 6 (TG PVM) (Fig. 1D). CD4⁺ T cells expressing IL-21 were enriched for expression of the ICOS/CXCR5 surface markers characteristic of Tfh cells (Fig. 1E, 1F) compared with CD4⁺ T cells that did not express IL-21, with a particularly high percentage of Tfh cells in mediastinal lymph node (MLN) IL-21 expressors.

The increased numbers of IL-21–producing CD4⁺ T cells in the lung after PVM infection suggested that this cytokine might participate in antiviral host defense and/or viral pathogenesis. To investigate this possibility, WT and Il21r^−/− mice were infected with PVM, and cellular inflammatory responses were evaluated. As observed previously (30, 31), lung tissue from PVM-infected WT mice exhibited progressive, severe inflammation diffusely through the lung at days 5 and 6, whereas lungs from Il21r^−/− mice had only mild and focal perivascular inflammation at these time points (Fig. 2A). Infiltration of both neutrophils (arrowheads) and lymphocytes (arrows) were evident in the WT mice (Fig. 2B). We also examined BAL fluid from PVM-infected mice and found significantly fewer cells in the BAL fluid from the Il21r^−/− mice than from WT mice on days 5, 6, and 9 postinoculation (Fig. 3A), as well as fewer cells in lung parenchyma at days 6 and 9 (Fig. 3B). Granulocytic infiltration is a hallmark of the early inflammatory response to PVM (28), and there were fewer Ly6G⁺CD11b⁺ neutrophils in the BAL fluid at days 5, 6, and 9 (Fig. 3C) and lung parenchyma (Fig. 3D) at day 6 in the Il21r^−/− than WT mice inoculated with PVM. Consistent with this, at day 6 postinoculation, Il21r^−/− mice also had significantly decreased levels of mRNAs encoding matrix metalloprotease-8 and S100A8 (Fig. 3E, 3F), two neutrophil-derived proteins involved in mediating the inflammatory/lung remodeling response (32, 33).

Because T cells were reported to be important for the resolution of sublethal PVM infection (34), we assessed the presence of CD4⁺ T cells, CD8⁺ T cells, γδ⁺ T cells, and NK cells in the lung during the course of PVM infection in WT and Il21r^−/− mice (Fig. 4). CD4⁺ T cell infiltration was apparent by day 6 after inoculation, but to a greater degree in WT mice than in Il21r^−/− mice, with elevated CD4⁺ T cell numbers persisting through day 9 postinfection (Fig. 4A). Increased numbers of CD8⁺ T cells were observed in the lungs of WT mice as early as 3 d after PVM inoculation but then declined, whereas fewer CD8⁺ T cells were observed in the lungs of infected Il21r^−/− mice (Fig. 4B). Although γδ⁺ T cells infiltrated both WT and Il21r^−/− lungs equivalently starting on day 5, fewer γδ⁺ T cells were seen in the Il21r^−/− lungs at days 6 and 9 (Fig. 4C). NK cell numbers were also significantly lower in the lung parenchyma in the Il21r^−/− mice (Fig. 4D). Thus, the recruitment of CD4⁺, CD8⁺, and γδ T cells, as well as NK cells, was diminished in the absence of IL-21 signaling.

CD8⁺ T cell function was previously reported to be suppressed in PVM-infected lungs (35). Interestingly, the percentage of lung CD8⁺ T cells producing IFN-γ increased during the course of infection in both WT and Il21r^−/− mice (Fig. 4E), but because of the reduced number of CD8⁺ T cells in the lungs of Il21r^−/− mice (Fig. 4B), the total number of IFN-γ–producing CD8⁺ T cells was significantly diminished (Fig. 4F). PVM peptides that allow the detection of viral-specific CD8⁺ T cells have not been identified; thus, these CD8⁺ T cells may include both PVM-specific cytotoxic cells as well as bystander cells activated by the inflammatory environment.

The inflammatory response to PVM in WT mice includes rapid infiltration of both neutrophils and lymphocytes into the lung, and, as noted above, these responses were reduced in Il21r^−/− mice. Neutrophil infiltration in PVM infection is coordinated by multiple cytokines and proinflammatory chemokines (28, 30, 31). In other settings, IL-17 has been shown to promote neutrophil responses, leading to the induction of cytokines and chemokines that augment the levels of neutrophil progenitors and the subsequent expansion of peripheral neutrophil populations (36). Because IL-21 promotes the differentiation of IL-17–producing cells (12—14), we measured the levels of cytokines associated with Th17 and Tc17 responses after PVM infection. Levels of both Il17a (Fig. 5A) and Il22 (Fig. 5B) mRNAs in lung tissue were not significantly different in WT and Il21r^−/− PVM-infected mice. Tnfa (Fig. 5C) and Ifng (Fig. 5D) mRNAs tended to be slightly lower in the Il21r^−/− mice at day 6 after inoculation, but the differences were not statistically significant. I11b mRNA was significantly decreased at days 6 and 9 (Fig. 5E), although differences in IL-1β protein were not observed (data not shown). However, Il6 mRNA was significantly decreased at days 5, 6, and 9 (Fig. 5F), with a corresponding decrease in IL-6 protein at day 6 in both BAL fluid and lung tissue (Fig. 5G, 5H). These results suggest that IL-21 signaling may directly or indirectly control the production of IL-6 during PVM infection.

Because the accumulation of neutrophils in the lungs was diminished in PVM-infected Il21r^−/− mice (Fig. 3), we examined the expression of several chemokines known to promote neutrophil recruitment during PVM infection (37). Interestingly, Cxcl1 mRNA was less potently induced in lung tissue of Il21r^−/− mice than in WT mice during PVM infection (Fig. 6A), whereas mRNA encoding CXCL10, a chemokine involved in lymphocyte recruitment, was induced with slightly more rapid kinetics in WT mice (Fig. 6B), but by day 6, the levels were not significantly different in WT versus Il21r^−/− mice. We observed no significant difference in the expression of the chemokine CCL3 (also known as MIP-1α), which has been implicated in neutrophil recruitment and immunomodulatory protection in PVM infection (28) (Fig. 6C), and although lower Cxcl2 mRNA expression was observed in Il21r^−/− mice at day 6 (Fig. 6D), there was no significant difference in CXCL2 protein expression in the lung (data now shown). Corresponding to the diminished levels of Cxcl1 mRNA (Fig. 6A), CXCL1 protein levels were significantly lower in BAL fluid (Fig. 6E) and lung homogenate (Fig. 6F) at days 5 and 6 after PVM infection of the Il21r^−/− mice, correlating with the reduced neutrophil infiltration in the lungs of these mice.

Because of the roles that IL-21 plays in the development of cellular immune responses, it was of interest to determine whether changes in cellularity, chemokines, and cytokines were specific to the lung response or whether they reflected an overall deficiency of the immune response in Il21r^−/− mice. We thus examined lymphoid populations in the MLN at days 0 and 6 after PVM infection and found that both WT and Il21r^−/− MLN undergo T cell expansion (Fig. 7A), including both CD4⁺ (Fig. 7B) and CD8⁺ (Fig. 7C) in response to infection, although this was reduced in the Il21r^−/− MLN (Fig. 7A–C). Similar levels of Ifng, Tnfa, and Il6 mRNAs were measured in the MLN, although Il17a and Il1b mRNA levels were lower in the Il21r^−/− MLN (Fig. 7D–H). These data indicate that although lymphoid expansion was lower in the Il21r^−/− MLN, most inflammatory cytokine responses were similar, suggesting that the cellular immune response was not completely defective in the knockout mice.

The production of IL-21 by CD4⁺ T cells in the normal lung suggested that IL-21 may play a role in normal lung homeostasis as well as in the development of the inflammatory response to PVM infection. To investigate this possibility, we examined lung cellularity in transgenic mice that constitutively express IL-21 in immune cells (TG21) (5). We observed increased cellularity in both the BAL fluid (Fig. 8A) and the lung (Fig. 8B), with an elevated percentage of neutrophils in both BAL and lung in the uninfected TG21 mice (Fig. 8C). Levels of Cxcl1 mRNA (Fig. 8D) and Il6 mRNA (Fig. 8E) were both increased in the TG21 lungs. Interestingly, however, levels of the neutrophil-expressed Mmp8 mRNA were similar in WT and TG21 lungs (Fig. 8F), suggesting that the enhanced recruitment of neutrophils to the lung occurred without an associated inflammatory response. To determine whether IL-21 could directly induce the production of IL-6, we purified DC populations and found that IL-21 significantly induced both IL-6 mRNA and protein (Fig. 8G, 8H) in these cells.

Above, we observed an increase in IL-21 production in response to PVM infection as well as decreased neutrophil and lymphocyte accumulation and diminished production of IL-6 in Il21r^−/− mice. It was unclear, however, whether IL-21 mediated the PVM pathogenic response or instead promoted host defense. To clarify the role of IL-21, we inoculated WT and Il21r^−/− mice intranasally with a dose of PVM previously shown to be sufficient to kill WT mice and determined their survival (Fig. 9A). As anticipated, 60% of the WT mice died on day 7, and all were dead by day 10 after PVM inoculation. In contrast, 90% of the Il21r^−/− mice were still alive at day 7, and 30% of the Il21r^−/− mice survived beyond day 10, with 5% (1 out of 19 mice) surviving through day 21. Thus, the absence of IL-21 signaling conferred a survival advantage (p < 0.0001), even though there was no significant difference in PVM SH gene expression, as an indicator of virus copy number, in WT versus Il21r^−/− lungs (Fig. 9B). These results suggest that IL-21 does not have a significant effect on viral replication or clearance but that it promotes the inflammatory response following infection with PVM, with earlier and augmented mortality in response to PVM infection in WT mice.

To determine whether the enhanced survival of the Il21r^−/− mice was a direct result of the effects of IL-21 in the lung during the response to PVM infection, we next used an IL-21R/Fc fusion protein to block IL-21 activity. When WT mice were intratracheally treated with the IL-21R/Fc fusion protein 1 d prior and 2 d after PVM inoculation, there was significantly higher survival than in mice treated with an Fc control protein (Fig. 9C), even though no significant differences in virus replication, based on SH gene expression, were detected between these two groups at day 6 of infection (Fig. 9D). Moreover, when we inoculated mice with a lower dose of PVM, treatment with the IL-21R/Fc fusion protein conferred complete protection (Fig. 9E), and again there were no differences in viral burden (Fig. 9F). These data indicate that the survival advantage seen in Il21r^−/− mice did not result from developmental differences but rather resulted from the lack of IL-21 signaling.

In this study, we have shown that IL-21 signaling plays a critical role in promoting the lung inflammatory response to acute pneumovirus infection. IL-21 expression in the lung in response to acute or chronic viral infections in general and to PVM in particular has not previously been reported. Our studies revealed that IL-21 mRNA levels are significantly elevated in lung tissue after PVM infection and that this induction parallels increased lung inflammation. Correspondingly, studies with Il21-mCherry reporter mice showed that enhanced IL-21 expression in the lung after PVM infection is specific for CD4⁺ T cells. Interestingly, uninfected lungs also contained a small but detectable population of CD4⁺ T cells expressing mCherry, suggesting a potential role for IL-21 in basal lung homeostatic responses to environmental Ags.

To determine if the production of IL-21 promoted or diminished PVM-induced disease, we examined PVM-inoculated Il21r^−/− mice and found they had significantly reduced lung pathology, with decreased infiltration of both neutrophils and lymphocytes. Previous studies showed that CCL3/MIP-1α, a chemokine produced by lung epithelium in response to RSV infection (38), was critical for neutrophil accumulation in the lungs and survival after PVM infection (39). Interestingly, CCL3 levels in the Il21r^−/− mice were at most modestly diminished, although neutrophil infiltration was significantly reduced. In contrast, the level of CXCL1, which also mediates neutrophil chemotaxis (40), was much less potently induced in the BAL fluid and lungs of Il21r^−/− mice than in WT mice; thus, the diminished CXCL1 may at least in part explain the diminished neutrophil accumulation in Il21r^−/− mice.

Our results reveal a potential role for IL-21 in normal lung homeostasis. Uninfected IL-21 transgenic mice had increased lung cellularity, including neutrophilia. Additionally, lungs from TG21 mice had increased levels of IL-6 and CXCL1, both involved in mediating inflammatory pathology. Interestingly, IL-21 could directly induce IL-6 expression in DCs, providing a direct link between IL-21 and the inflammatory process. Previously IL-6 was shown to induce IL-21 expression in CD4⁺ T cells (14, 41). Our results suggest that IL-21 can potentially amplify its own induction via the induction of IL-6.

T cell responses significantly contribute to the host response to sublethal PVM infection, with both CD4⁺ and CD8⁺ T cell subsets contributing to viral clearance (34). Although both CD4⁺ and CD8⁺ T cell numbers were decreased in the Il21r^−/− mice, there was no apparent change in viral replication or clearance. Th17 cytokines are known to mediate inflammation, and IL-21 serves to promote Th17 differentiation (12–14); however, there was no apparent decrease in Th17 cells or of IL-17A or IL-22 production in the lungs of PVM-infected Il21r^−/− mice. Thus, Th17 differentiation does not appear to play a substantial role in PVM infection.

It was previously reported that most PVM-specific CD8⁺ T cells produce diminished levels of cytokines upon antigenic stimulation (35). In our studies, however, there was no difference in the percentage of CD8⁺ T cells that could produce IFN-γ in WT and Il21r^−/− mice. However, the total number of CD8⁺ T cells was greatly decreased, consistent with the known role of IL-21 for CD8⁺ T cell expansion and differentiation (4, 42), and thus the absolute number of IFN-γ⁺CD8⁺ T cells was also greatly reduced in the Il21r^−/− lungs. It is possible that although the absence of IL-21 did not affect viral clearance in the primary response to PVM, CD8⁺ T cells primed in the absence of IL-21 might have defective memory responses to PVM upon secondary challenge.

In summary, we have demonstrated that IL-21 plays a previously unappreciated role in the host response to PVM, promoting an unanticipated pathogenic inflammatory response to the virus that is associated with reduced host survival, which is in marked contrast to the protective role played by IL-21 in chronic lymphocytic choriomeningitis virus infections and its correlation with control of HIV. Importantly, we also found that pretreatment of WT animals with a soluble IL-21R–Fc fusion protein was protective, an observation with therapeutic implications for PVM and potentially for other viral infections as well.

We thank Caroline M. Percopo, National Institute of Allergy and Infectious Diseases, for technical assistance and Dr. Helene F. Rosenberg, National Institute of Allergy and Infectious Diseases, for critical comments/valuable discussions, inoculating some of the mice with PVM, and providing PVM. We also thank Dr. Derry Roopenian and Thomas Sproule, The Jackson Laboratory, for breeding some of the Il21r^−/− mice used in this study and Dr. Jian-Xin Lin, National Heart, Lung, and Blood Institute, for critical comments.

W.J.L. and R.S. are inventors on patents and patent applications related to IL-21. The other authors have no financial conflicts of interest.